DOGE used ChatGPT in a way that was both dumb and illegal, judge rules

DOGE used ChatGPT in a way that was both dumb and illegal, judge rules

法官裁定:DOGE 使用 ChatGPT 的方式既愚蠢又违法

The Department of Government Efficiency’s cancellation of over $100 million in grants was unconstitutional, according to a ruling on Thursday. In the 143-page decision, US District Judge Colleen McMahon cites DOGE’s process for eliminating grants, which involved using ChatGPT to determine if something is related to diversity, equity, and inclusion (DEI).

根据周四的一项裁决,政府效率部(DOGE)取消超过 1 亿美元拨款的行为违宪。在美国地方法官科琳·麦克马洪(Colleen McMahon)长达 143 页的判决书中,她指出了 DOGE 在取消拨款时的流程——即利用 ChatGPT 来判定项目是否与“多元、公平和包容”(DEI)相关。

The ruling, which stems from a 2025 lawsuit filed by humanities groups, says “it could not be more obvious that DOGE used the mere presence of particular, protected characteristics to disqualify grants from continued funding” from the National Endowment for the Humanities (NEH). Judge McMahon cites several instances in which DOGE appeared to use ChatGPT to scan and eliminate grants using their relation to characteristics like race, national origin, religion, and sexuality.

该裁决源于人文团体在 2025 年提起的一项诉讼。判决书称,“DOGE 仅仅因为项目中存在特定的受保护特征,就取消了国家人文基金会(NEH)的持续资助,这一点再明显不过了。”麦克马洪法官列举了多个案例,显示 DOGE 似乎利用 ChatGPT 扫描并剔除那些涉及种族、国籍、宗教和性取向等特征的拨款项目。

The filing mentions testimony from Justin Fox, a DOGE staffer who worked with his colleague Nate Cavanaugh to eliminate 97 percent of grants under the NEH, in part by relying on ChatGPT’s understanding of DEI:

法庭文件提到了 DOGE 工作人员贾斯汀·福克斯(Justin Fox)的证词。他与同事内特·卡瓦诺(Nate Cavanaugh)合作,通过依赖 ChatGPT 对 DEI 的理解,取消了 NEH 旗下 97% 的拨款项目:

Fox testified that he used ChatGPT “[t]o highlight why [a] grant may relate to DEI” and “to pull out anything related to DEI.” To do so, he submitted each cursory grant description from the NEH spreadsheet to ChatGPT using a standardized prompt: “Does the following relate at all to DEI? Respond factually in less than 120 characters. Begin with ‘Yes.’ or ‘No.’ followed by a brief explanation.” Fox testified that he did not define “DEI” for ChatGPT and that he did not have the slightest idea how ChatGPT understood the term.

福克斯作证称,他使用 ChatGPT 是为了“突出某项拨款为何可能与 DEI 相关”以及“剔除任何与 DEI 相关的内容”。为此,他将 NEH 电子表格中每一项简略的拨款描述提交给 ChatGPT,并使用了一个标准化的提示词:“以下内容是否与 DEI 有任何关联?请在 120 个字符内做出事实性回答。以‘Yes.’或‘No.’开头,并附上简短解释。”福克斯承认,他从未向 ChatGPT 定义过什么是“DEI”,也完全不知道 ChatGPT 是如何理解这个术语的。

In addition to asking ChatGPT for signs that something is related to DEI, Fox also asked the AI chatbot to scan NEH grants for what he called “Detection Codes” related to “protected characteristics,” according to the filing:

根据文件显示,除了询问 ChatGPT 是否存在与 DEI 相关的迹象外,福克斯还要求该 AI 聊天机器人扫描 NEH 的拨款项目,以寻找他所谓的与“受保护特征”相关的“检测代码”:

After being deployed from DOGE to NEH, Justin Fox used search terms, which he labeled as “Detection Codes,” to identify grants that he dubbed the “Craziest Grants” and “Other Bad Grants.” The search terms included, among other terms, “BIPOC (Black, Indigenous, People of Color),” “Minorities,” “Native,” “Tribal,” “Indigenous,” “Immigrant,” “LGBTQ,” “Homosexual,” and “Gay.” When Fox was asked whether he “r[a]n this list of words through every grant description” he received from NEH, he confirmed, “yes.” In this way, Fox constructed and applied explicit classifications based on protected characteristics and used them as the operative criteria for revoking federal grants.

在从 DOGE 被派往 NEH 后,贾斯汀·福克斯使用了他标记为“检测代码”的搜索词,来识别他所谓的“最疯狂的拨款”和“其他糟糕的拨款”。这些搜索词包括“BIPOC(黑人、原住民、有色人种)”、“少数群体”、“原住民”、“部落”、“土著”、“移民”、“LGBTQ”、“同性恋”等。当被问及是否将这份词汇表“运行在从 NEH 收到的每一份拨款描述上”时,福克斯确认道:“是的。”通过这种方式,福克斯构建并应用了基于受保护特征的明确分类,并将它们作为撤销联邦拨款的执行标准。

Judge McMahon writes that DOGE deemed hundreds of grants “wasteful because they related to Blacks, women, Jews, Asian Americans, and Indigenous people,” adding that “the very subjects DOGE treated as markers of waste, lack of merit, or ideological contamination are the subjects that Congress made expressly germane to NEH’s mission.” Some of the grants lumped into the “wasteful” category related to projects about the Holocaust, civil rights, and an educational experience that would allow participants to “explor[e] indigenous knowledge, culture, and climate.”

麦克马洪法官写道,DOGE 将数百项拨款视为“浪费,因为它们涉及黑人、女性、犹太人、亚裔美国人和原住民”,并补充说,“DOGE 将其视为浪费、缺乏价值或意识形态污染的标志,恰恰是国会明确规定与 NEH 使命相关的主题。”一些被归入“浪费”类别的项目涉及大屠杀、民权,以及一项旨在让参与者“探索原住民知识、文化和气候”的教育体验。

McMahon also pushes back on the government’s argument that “there is no real constitutional problem here because any viewpoint-based classification was ChatGPT’s doing, rather than the Government’s:”

麦克马洪还驳斥了政府的辩解,即“这里不存在真正的宪法问题,因为任何基于观点的分类都是 ChatGPT 的行为,而非政府的行为”:

There is no distinction to be drawn here between the Government and ChatGPT. ChatGPT was the Government’s chosen instrument for purposes of this project, and DOGE’s use of AI to identify DEI-related material neither excuses presumptively unconstitutional conduct nor gives the Government carte blanche to engage in it. …There is not a scintilla of evidence that Fox or Cavanaugh, having obtained a “DEI” rationale from ChatGPT, undertook any meaningful review of whether that rationale made sense.

“政府和 ChatGPT 之间没有区别。ChatGPT 是政府为本项目选择的工具,DOGE 使用 AI 来识别 DEI 相关材料,既不能为推定违宪的行为开脱,也不能赋予政府随意进行此类行为的特权……没有任何证据表明,福克斯或卡瓦诺在从 ChatGPT 获得‘DEI’理由后,对该理由是否合理进行了任何有意义的审查。”

In her decision, Judge McMahon ultimately found that DOGE’s elimination of over 1,400 NEH grants was unlawful and unconstitutional, citing violations of the First Amendment, the Fifth Amendment’s equal protection law, and DOGE’s lack of authority. McMahon issued an order to reverse their cancellation.

在判决中,麦克马洪法官最终裁定,DOGE 取消 1400 多项 NEH 拨款的行为是非法且违宪的,理由是其违反了宪法第一修正案、第五修正案的平等保护条款,且 DOGE 本身缺乏相关授权。麦克马洪下达了撤销其取消决定的命令。